Loading timeline…
20122010s
AlexNet
The deep CNN that sparked the deep learning revolution by crushing the ImageNet benchmark.
Why It Was Important
Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton designed a deep convolutional neural network trained on GPUs that reduced the ImageNet top-5 error rate from 26% to 15.3%. This massive margin officially proved that deep learning methodologies drastically outperformed hand-coded feature extraction, triggering an industry-wide pivot.
Who Invented It
Alex Krizhevsky, Ilya Sutskever, Geoffrey Hinton
University of Toronto researchers who proved deep learning at massive scale.
Applications
- Computer Vision Revolution
- GPU Training Paradigm
- Image Classification
Key Papers
- ImageNet Classification with Deep Convolutional Neural Networks
Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · NeurIPS 2012
Videos
The moment we stopped understanding AI [AlexNet]
Welch Labs